mixedrealitylabfandomcom-20200215-history
Report for ICMI 2010, Beijing China - James Teh
12th International Conference on Multimodal Interfaces and 7th Workshop on Machine Learning for Multimodal Interaction Date: November 8-10, 2010 Venue: Beijing, China This is the first participation by MXR lab members in the ACM-SIGCHI ICMI conference. This conference is partly sponsored by NSF, Google and Microsoft. The conference is rather technical and very specific in contribution, and presentations are mostly focused on the sessions that they are presenting in. For example, the conference has a session on Gaze and Interaction where the presenters all presented about some aspects of visual attention and how we could apply visual analysis in applications such as Human-Robot Interaction, Gaming (Tetris) and Interactive Story Telling. For lab researchers involved in Robotics and Intelligent Agents, there are a few papers which might be of interest: Real-Time Adaptive Behaviors in Multimodal Human-Avatar Interactions Authors: Hui Zhang, Damian Fricker, Thomas G. Smith, Chen Yu '' Focusing Computational Visual Attention in Multi-Modal Human-Robot Interaction ''Authors: Boris Schauerte, Gernot A. Fink '' Recommendation from Robots in a Real-world Retail Shop ''Authors: Koji Kamei, Kazuhiko Shinozawa, Tetsushi Ikeda, Akira Utsumi, Takahiro Miyashita, Norihiro Hagita The more interesting paper presentation sessions are: ' Multimodal Systems The papers presented in this session mostly measure the performance of multimodal systems. They may focus on different modes of the communication channel, for example, one paper by Dan Bohus focuses on multiparty dialog with gaze, gesture and speech analysis and evaluation. This research reminds me a little of Dr Newton’s PhD work, so might be of interest to him. Gaze and Interaction The most interesting paper presentation for me in Day 1 is the paper Focusing Computational Visual Attention in Multi-Modal Human-Robot Interaction ''by Boris Schauerte, Gernot A. Fink. Authors showed how intelligent agents or robots can correctly identify referred to objects by combining, visual, gesture and voice information. Demo Session The demo session is small, with only 4 projects on showcase. Personally I felt that the Poster session on the last day was more interesting that the Demo session, and many conference attendees also felt the same way. Interestingly, there are 3 out of 4 projects that demonstrate some kind of speech recognition as input to a system for search and input purposes. For me the interesting demo is the Cloud Mouse, by researchers from Microsoft in Beijing and Cambridge. The nicely designed mouse, and new way of interaction by grasping and squeezing works well with maneuvering a large information display. The cloud mouse uses a “squeeze and direct point” mode instead of “translation and click” mode to control, which provides a more natural drag & drop experience with tangible feedbacks like touch and sound. Multimodal Interfaces This session has some pretty interesting projects. “Recommendation from Robots in a Real-world Retail Shop” sees the researchers putting physical networked robots in a retail shop and together with some digital displays. These robots observed, interact and analyze data from the shoppers that were in the shop and use the new knowledge to try and influence shoppers’ behavior. “3D-Press: Haptic Illusion of Compliance when Pressing on a Rigid Surface” is interesting for me. The authors show the idea of creating an illusion where users feel that the surface they are applying force is actually bendable even though it is rigid. It makes use of vibrotactile and pressure sensing technologies. '''Final Day Keynote speaker Prof James Landay gave an interesting talk titled “Activity-based UbiComp: A New Research Basis for the Future of Human-Computer Interaction”. He showed a variety of projects that his research group had done. There is also a focus on family and elderly (might be of interest to some of our Lab researchers). He gave a vision of how Mobile computing and Ubicomp will shape the future. From my interaction with him during lunch, he will be visiting NUS and to come to our lab for a visit. Other activities The Banquet Dinner on the second evening of the conference is one of the highlights of the conference. The venue is quite a historical site, because it was a home to one of the former Chinese royalty. Picture shows conference participants at the Dinner venue